Preparing Rackspace Data Centers for Hurricane Sandy(rackspace.com)
rackspace.com
Preparing Rackspace Data Centers for Hurricane Sandy
http://www.rackspace.com/blog/preparing-our-data-centers-for-hurricane-sandy/
6 comments
Makes sense - assuming supply lines are cut then you'll need to get in before the thundering herd clobbers access to fuel and jacks the price up to uneconomic short term levels. I'm surprised they don't keep large strategic oil reserves on site like the military does.
As it turns out I know the answer to that one. Its a question of hazardous materials. You need a permit to store fuel on site (and you have to get regular inspections to insure it isn't leaking into the groundwater) and the permits get more expensive the more fuel you keep. At some point the permitting authority would probably change your business classification from 'data center' to 'fuel depot' if you went overboard :-)
I believe some do keep strategic reserves, they just want to "reserve" their usage for situations where they don't have other options.
It is nice that they talk about this - however these are not Rackspace standards but standard industrywide, some of which have been in place since the 1960s, when ATT revised their plans for central office facilities in the wake of riots in some American cities.
It is frustrating for customers outside of the US to have to look up and convert these unfamiliar time zones. As a company that services customers all over the planet, why not post UTC?
Thankfully though, they managed to convert the generator run times from the completely unfamiliar "hours" unit into the much more universal "days" unit. </sarcasm>
Thankfully though, they managed to convert the generator run times from the completely unfamiliar "hours" unit into the much more universal "days" unit. </sarcasm>
This link returns a 404 error for me.
Sweet, sweet irony. http://cl.ly/image/2G1Z1A3c1S0c
I hope preparing didn't mean putting up a 404 page. =)
But given the comments above, it seems likely that the issue wasn't a result of lack of power/gas. Interested to find out what failed with the preparations.
But given the comments above, it seems likely that the issue wasn't a result of lack of power/gas. Interested to find out what failed with the preparations.
Seeing reports of data centers experiencing flooding I really wonder what best practices are for a data center in a flood zone. Not like you could float it or anything. It hasn't been on my list of worries (Santa Clara doesn't flood much, more likely to fall into a crack in the ground). Anyone here who has an action plan for flooding at their data center? I recall seeing a nuclear plant that had a levee wall, that was pretty impressive.